Valuable data and analysis results obtained through surveys cannot fully realize their worth if not communicated properly. Data visualization is a powerful tool that transforms complex and vast survey results into an easily understandable and memorable format, effectively conveying key messages and supporting data-driven decision-making. A single well-made chart can be more persuasive than countless numbers.
However, simply transferring data to a chart is not enough. Incorrect chart selection or inappropriate design can distort information or mislead viewers. For example, a pie chart is suitable for showing the proportion of parts that make up a whole 59, but a line graph is much more effective for showing changes over time. If time-series data is represented by a pie chart, it becomes difficult to identify trends and can lead to misinterpretation.
Conditions for Good Data Visualization:
- Clarity: The key information to be conveyed should be clearly visible at a glance. Simplicity is more important than complexity.
- Accuracy: Data must be represented honestly and accurately without distortion. Expressions that could cause misunderstanding should be avoided.
- Simplicity: Minimize unnecessary decorations, colors, text, etc., and design to focus on the data itself.
- Context Provision: Clearly present chart titles, axis labels, legends, data sources, etc., to help understand the meaning and background of the data.
- Storytelling: It should not just list data but convey a meaningful message or story through the data. Visual elements should be used to elicit understanding and empathy from the viewer.
Guide to Selecting Appropriate Charts Based on Data Type and Communication Purpose:
The following are common chart types used in presenting survey results and their applications.
- Proportion/Composition Comparison (Parts of a Whole)
- Example:
- Proportion of respondents selecting each option (gender, satisfaction rating) out of the total
- Recommended Chart Type:
- Pie Chart
- Donut Chart
- 100% Stacked Bar Chart
- Advantages:
- Intuitively shows the proportion of each part to the whole.
- Visually easy to understand.
- Disadvantages & Points to Note When Using:
- Readability significantly drops if the number of items is too large (generally more than 5).
- Difficult to compare subtle differences between items.
- 3D pie charts should never be used as they cause visual distortion.
- Good Visualization Example (Brief Description):
- Displaying the response ratio for "Most Preferred SNS Channel" as a pie chart (e.g., Instagram 40%, YouTube 30%, Facebook 20%, Other 10%)
- Example:
- Item Comparison
- Example:
- Comparison of satisfaction scores by product, comparison of purchase frequency by age group
- Recommended Chart Type:
- Bar Chart (Vertical/Horizontal)
- Advantages:
- Clearly compares the size of each item.
- Horizontal bar charts are useful when item names are long.
- Disadvantages & Points to Note When Using:
- Can look complicated if the number of items is too large.
- The Y-axis (value axis) must start at 0 to prevent data distortion.
- Good Visualization Example (Brief Description):
- Comparing "Average Customer Satisfaction Scores for Competitor A, B, C and Own Product" using a vertical bar chart
- Example:
- Change/Trend Over Time
- Example:
- Monthly website visitor count change, quarterly product sales trend
- Recommended Chart Type:
- Line Graph
- Area Chart
- Advantages:
- Effectively shows data change trends over time.
- Easy to compare by displaying multiple data series together.
- Disadvantages & Points to Note When Using:
- Can look complicated if there are too many data points or high volatility.
- Trends can appear exaggerated or minimized depending on Y-axis scale settings.
- Good Visualization Example (Brief Description):
- Displaying "Monthly Customer Inquiry Count Trend Over the Past Year" as a line graph
- Example:
- Relationship/Correlation
- Example:
- Relationship between advertising expenditure and sales revenue, relationship between age and purchase intent for a specific product
- Recommended Chart Type:
- Scatter Plot
- Bubble Chart
- Advantages:
- Visually easy to identify relationship patterns between two variables (positive correlation, negative correlation, no correlation).
- Outliers can be easily detected.
- Disadvantages & Points to Note When Using:
- Difficult to identify relationships if data points are too few.
- Must note that correlation does not imply causation.
- Good Visualization Example (Brief Description):
- Displaying "Customer Age (X-axis) and Average Monthly Purchase Amount (Y-axis)" as a scatter plot to analyze consumption patterns by age group
- Example:
- Frequency Distribution
- Example:
- Distribution of respondent counts for specific score ranges (test scores, satisfaction scores)
- Recommended Chart Type:
- Histogram
- Advantages:
- Useful for identifying the distribution shape of continuous data (normal distribution, skewed distribution, etc.).
- Visually shows concentrated and dispersed data intervals.
- Disadvantages & Points to Note When Using:
- The shape of the graph can change depending on the number or width of bins, requiring careful interpretation.
- Good Visualization Example (Brief Description):
- Displaying "Distribution of Customer Satisfaction Survey Response Scores (1 to 10 points)" as a histogram
- Example:
- Geographic Data Visualization
- Example:
- Product sales by region, proportion of specific opinions by city/province
- Recommended Chart Type:
- Choropleth Map
- Advantages:
- Shows the distribution or pattern of data according to geographical location at a glance.
- Intuitive as it expresses value size by color intensity or type.
- Disadvantages & Points to Note When Using:
- Visual illusion can occur depending on the size of administrative districts (larger areas may appear more important).
- Using too many colors can cause confusion.
- Good Visualization Example (Brief Description):
- Displaying "National City/Province-wise Brand Awareness Level of Our Product" as a choropleth map (higher awareness represented by darker colors)
- Example:
- Open-Ended Response (Subjective Answer) Visualization
- Example:
- Keywords for customer complaints, summary of product improvement ideas
- Recommended Chart Type:
- Word Cloud
- Advantages:
- Visually emphasizes frequently mentioned words or phrases in subjective responses, allowing quick grasp of key content.
- Useful for showing the overall tendency of text data.
- Disadvantages & Points to Note When Using:
- Only reflects word frequency, making it difficult to express context or depth of emotion.
- Prior data cleaning, such as removing meaningless words (particles, conjunctions), is important.
- Good Visualization Example (Brief Description):
- Displaying "Most Frequently Mentioned Positive/Negative Keywords in Product Usage Reviews" separately as word clouds
- Example:
Common Data Visualization Mistakes and Improvement Measures:
- Incorrect Chart Type Selection: Using a chart that doesn't match the message or data characteristics, distorting information or making it difficult to understand.
- Bad Example: Using a pie chart to show changes over time.
- Improvement Measure: Clearly define the nature of the data (categorical, continuous, time-series, etc.) and the communication purpose (comparison, trend, distribution, etc.), and select the optimal chart type for each purpose. (Refer to the table above)
- Excessive Information or Unnecessary Decoration (Chart Junk): Displaying too many data series at once, or using unnecessary 3D effects, shadows, flashy colors, complex background images, etc., which prevent the important data from being noticeable.
- Bad Example: Representing monthly sales of multiple products with a complex 3D stacked bar chart, making it difficult to grasp the trend for each product.
- Improvement Measure: Focus on the key message to be conveyed and boldly remove unnecessary visual elements. Use simple and clean 2D charts as a basis, and use colors or effects sparingly only for parts that need emphasis.
- Distorted Axis Usage (Misleading Axes): Not starting the Y-axis (value axis) at 0, or setting axis intervals unevenly, thereby exaggerating or minimizing the actual differences or changes in data.62 This can lead to serious information distortion, whether intentional or not.
- Bad Example: Product A and B satisfaction scores are 3.5 and 3.8 respectively, but the Y-axis starts from 3.0, making the bar length difference appear very large.
- Improvement Measure: For bar charts or line graphs, ensure the value axis always starts at 0, and set axis intervals consistently to appropriately represent the entire data range.
- Inappropriate or Confusing Color Usage (Poor Color Choices): Using too many colors, using flashy colors without meaning, or using color combinations with low contrast against the background, reducing readability.
- Bad Example: A bar chart representing 10 categories with completely different intense primary colors, making it visually tiring and difficult to grasp the core.
- Improvement Measure: Use colors consistently and meaningfully. Use noticeable colors for key emphasis items, and use similar shades or achromatic colors for other items to create a visual hierarchy.61 Consistently using brand colors is also a good method.
- Missing or Unclear Labels, Legends, Titles: Lack of information clearly indicating what each element of the chart means, what unit it was measured in, or what the entire chart is about.
- Bad Example: A line graph showing only numbers without any explanation for the axes.
- Improvement Measure: Attach clear and concise titles to all charts, and specify what each axis represents along with units. If using multiple data series, always include a legend to distinguish each series.
Effective data visualization goes beyond merely making things look good; it is a key communication skill that clearly conveys insights gained from survey results and helps viewers make correct data-based judgments. Marketers must accurately understand the message they want to convey and the characteristics of the data, and select the most appropriate visualization method to maximize the value of survey results.